The keypad worked flawlessly, though. Also, often times when you put a battery in this thing, it turns on by itself with garbage on the screen. It doesn't even do that any more.

While I had it apart I used a magnifier and checked all the solder joints - everything looked perfect. Perhaps I will re-flow them anyway some day, for the heck of it, but I really can't think of any other reason it would go from working perfect to completely dead overnight besides static.

Cool! Like RPN before, I have no clue how slide rules work, so I've never had in an interest in them. That would probably change if I got a hold of one and was ambitious enough to learn how to use it. 

These rulers are based on the two basic laws about logarithms:
ln(A*B) = ln(A) + ln(B)
and (partially related to above):
ln(1/A) = -ln(A)

The sum is realized by sliding the segments (one centimeter on one slider plus two on the other one means three cm in total), the ln(X) is there printed on the sliders, on one of them starting from both sides (so allowing the "1/X") and on the other usually with one or two linear scales, so you can directly read there ln(X), e^X, log(X) and 10^X.

So if you want to multiply two numbers (e.g. 165 * 2.7), you first rewrite them into the "scientific" form (so 165 = 1.65E2; 270 = 2.7E0).
Exponents you sum up directly (that you can do in your head after just a short practice).
Then you set the rulers so, the "1" of the sliding ruler is aligned to the representation of the first number (1.65) on the main body
Then move the slider so, the hair marker is aligned with the second number on the sliding ruler (2.7).
Then that slider hairline shows the multiplication result (something around 4.45).
Then you "put back" the sum of the exponents and get the final result(so 445; correct result is 445.5)
Well, this was the basic example,

There were many modifications and extra features appearing on some models, like double slider, extra scales with special functions (sin, cos, tan,...), two sliding rulers, larger models for better accuracy, smaller models for compact size, special scales for some specific tasks (navigation calculations,...) and so on...

It is of course by far not that accurate as the digital calculators, but for the quick engineering calculations it was way sufficient. With the knowledge of the concept of logarithms and a bit of practice you can do most of the calculations and do them really quick, I would say even quicker than with the electronic one. Of course, the lack of functions (you need a table for that), accuracy and the sensitivity to manipulation error are the disadvantages, but it never has flat battery problem...

I just bought my replacement for the dead TI-1025. I got a perfect working, mint condition Canon Multi 8 for $20 on eBay, a calculator which I've wanted for several years now. The Canon Multi 8 is a very unique calculator in that it was the first handheld calculator to have a dual-line display, and the only calculator to have such a VFD display. Dual-line displays wouldn't be seen again until LCD technology made them more practical to manufacture.

The Multi 8 can be run in three modes; a single-line mode in which it functions as a normal single-line calculator, a dual-line mode in which when performing a calculation on two numbers, you can see both numbers, and the operation symbol, before pressing equals; and a memory display mode in which it operates like a single-line calculator, but the upper line always displays what's in memory.

As unique as it is, the Multi 8 is a surprisingly cheap and easy calculator to find. Not common by any means, but not rare. In my opinion, out of the four American and Japanese calculator manufacturers, Canon's units are the most stylish. The Palmtronic 8M is a good example with its color scheme. Compare this to the bland aesthetics (and bulk) of something like the TI-1025.


Here's what I bought. I probably won't get it in time for my Digital Systems exam, in which case I'll have Mom put it away for Christmas.

http://www.ebay.com/itm/161484042787

Whatever you do, DO NOT put it in your jacket pocket!

lol, Nope. Not making that mistake again.

I got the HP-33E today.  Works absolutely perfect, and in excellent shape. I *was* going to solder a 2x AAA battery holder onto it, but as it turned out, the one intact battery contact fell off as soon as I touched it, so now two AA batteries fit in and work perfectly. I like it a lot, I've used it quite a bit.

Very cool! Just make sure to be extremely careful putting in batteries. I've seen cases so bad that even the wiring fell completely apart.

Oh, the contacts aren't rotten or anything. Back in the day people would try and use regular AA batteries, which are slightly longer than the original battery pack, and so would fit very tightly. The bending of the battery contacts under the physical stress would cause them to break off. A small piece of the contact still remains, though, so once the contacts have broken off, regular AA's fit in perfectly.

I've used this thing a ton, and I like it a lot. Hasn't missed a beat. Even completed an entire Electric Circuits assignment with it. I kinda feel bad because Mom left it here with me instead of taking it home and wrapping it up for Christmas, because I said I wouldn't use it, which would have been true had the remaining battery contact not broken off, allowing me to use AA batteries. Oh well.  

Although I'm glad, I'm also kind of surprised this thing works perfectly. These particular HP calculators suffered a serious design flaw which would cause them to stop working even while still under warranty, let alone 35 years later. The electronics inside the calculator aren't actually soldered - instead the circuit board is a sheet with metal contacts and traces printed on it, and the components are simply laid on it, and held to it via a plastic guide placing pressure on them. It was a poor decision in an attempt to reduce manufacturing costs, and it proved so unreliable that HP switched back to the conventional PCB/soldered design in 1983. Mine was made in 1978, and so presumably uses the former design. But it's worked so perfectly in all my use of it so far, so I wonder if it might have actually been sent to HP for repair some time in its life, and had the guts replaced with the soldered design. Who knows.

I thought I would never make use of the programming on this (I would have originally accepted a 32E over a 33E since the 32E has a few more features at the expense of programming), but I've actually already written a couple of programs for this. One calculates pi using Leibniz's formula, and it pauses after each iteration to show the result. It gets it to one decimal place (3.1) in a minute, two decimal places (3.14) in 21 minutes, and 3 decimal places in an hour and a half. Of course those times are limited by the fact that it pauses. I also wrote an alternate version that doesn't pause, which would be much faster, although I haven't timed it yet. I ran that version for 19 hours until the batteries died, and it still didn't get to 4 decimal places. You can watch it running the version of the program that displays each iteration result here:

https://www.youtube.com/watch?v=Qz7XhGPNGE8

Edit: Corrected pi calculation times.

Cool! How long did you enter and debug the program? Must be pretty tough to program such an old calculator, isn't it?
But you have a display here, at least. When setting up old 9pin matrix printers you got each response printed on a separate sheet.

Oh no, it's super easy to program this. All a program consists of is keystrokes that the calculator runs through in succession. For example, if I wanted a program that calculates 4*5, I just put it in program mode and press 4, Enter, 5, * (just as if you wanted to calculate it normally), and that's the program. When I run it, it will give me 20. You can also do basic loops and If statements, though.

The pi calculation is 27 lines of "code" (this calculator has room for 49), and can be entered in about 30 seconds. It took me maybe half an hour to figure out the algorithm. Most keystrokes make up one line, but some operations that require two or three keystrokes (such as storing a value in memory) can be concatenated into one line.

Ah, English please???   

What part don't you get? 

The part that has to do with math... What in God's name are Polar Numbers? Are they numbers so cold that they must be heated up with complicated math before they will work? 

I can't find where I've even mentioned polar numbers here! Unless you mean from my video.  

Anyway, there's two ways of representing complex numbers - rectangular and polar. Rectangular is just the standard way of plotting stuff like you've probably already done. An example of a complex number in rectangular form is 3+4i. 3 on the x-axis, and 4 on the y-axis. The i means the imaginary component of the number (complex numbers have a "real" and "imaginary" component). In Electrical Engineering we use j instead of i since we already use I for electrical current.

In polar form, instead of representing the number with an x-direction and y-direction, we use a length, and an angle above the x-axis. 3+4i in polar form is 5 angle 53.13 degrees. You can do Pythagorean theorem to get the length (sqrt(3^2 + 4^2) = 5), and then trigonometry to get the angle (tan^-1 (4/3) = 53.13).

Numbers are fun!